1

Process Intensification in Industrial Wastewater Treatment Gary Howard, Principal Consultant Process Engineer, Foster Wheeler, Reading, UK

gary_howard@fwuk.fwc.com Keywords: wastewater, treatment, industrial, zero, effluent, discharge, intensification, footprint, WM-SET (Water Management – Site Evaluation Tool) Abstract In recent years there has been an increased focus on the environmental impacts of process plants (oil and gas, refineries, petrochemical and pharmaceuticals). Tightening global environmental standards have required new and existing facilities to achieve higher standards of effluent treatment and to recover more water for re-use. Economic pressures require schemes to be as energy efficient as possible and to use minimal space. Process intensification can help address these issues. This paper looks at drivers for intensification in wastewater treatment and water recovery. It reviews typical intensified effluent treatment technologies and presents some recent examples where intensification opportunities have been considered. It also examines the challenges faced when dealing with intensified processes and the steps needed to successfully manage intensification. Drivers for Intensification Environmental standards are continuing to become more onerous. Consents have addressed macro-pollutants and these have been eliminated or reduced to a benign level in treated wastewater. Regulators are now targeting minor components in industrial effluents. In some countries environmental standards are so strict that effluent treatment systems are being driven towards Zero Effluent Discharge (ZD) systems. To compound this, in many areas of the world there is a deficiency in water availability and ZD is the inevitable regulatory requirement. In other countries, standards are not yet as stringent, but where a project is being financed by international banks, export credit agencies or multilateral agencies, it will be required to meet the World Bank/International Finance Corporation (IFC) Performance Standards. Process plants require considerable quantities of water for both processing and evaporative cooling. A typical refinery uses about 2.5 m³ of water per tonne of product; about 60 % of that requirement is lost as evaporation to achieve cooling. In context, a typical 400,000 barrels per year refinery uses approximately enough water volume every hour to fill the equivalent of an Olympic-sized swimming pool. For both greenfield and brownfield developments the technology needed to achieve the more stringent discharge standards and the required water recovery efficiencies has, of necessity, become more complex. These more complex systems also use a larger plot space which, in most modern process plant sites, can be at a premium. The dual pressure of additional unit operations and a drive to use less space makes process intensification a serious consideration. Reducing energy usage is another key objective of industrial facilities. Within effluent treatment facilities the easiest energy savings are typically achieved by using more efficient motors and oxygen transfer equipment. However, the increasing complexity of the required equipment invariably means that, overall, the energy required to treat a cubic metre of effluent tends to increase rather than decrease. Technologies This paper reviews typical intensification options for two common wastewater treatment stages,

namely, oil removal and biological treatment. The location of these stages within a typical ZD scheme

is shown in Figure1.

2

Figure 1 - Typical ZD Scheme

Raw water is treated to allow use in boilers, cooling towers, industrial process units and for other purposes. Oily water is pre-treated to remove oily components that would adversely affect a biological effluent treatment plant and is cooled to a suitable temperature. The de-oiled effluent is balanced and treated in a biological treatment plant to remove biodegradable compounds. The biological process generates sludge which is removed and thickened for disposal, typically off-site. Following biological treatment, the dissolved solids need to be reduced or removed for most re-use applications. This is usually achieved by either Reverse Osmosis (RO) or Multiple Effect Distillation (MED). This water-upgrading step results in a concentrate which can be further reduced by solar evaporation in arid climates or by using evaporator/crystalliser systems in temperate climates or where space is at a premium. Both of these processes produce a slurry or solid which will ultimately require disposal, typically in a controlled landfill. Oil Removal

The oil removal technologies commonly used in industrial effluent treatment systems are:

• American Petroleum Institute (API) separators • Corrugated Plate Interceptors (CPI) • Dissolved Air Flotation (DAF)

Other technologies which have been used for specific purposes include coalescers, filters, membranes and centrifuges.

3

API Separator Oil separators designed according to the API guidance are simple gravity separators which rely upon density differences to separate free-phase oil droplets and solids. For an API separator, Stokes’ law is used, with safety factors, to develop a design which will remove oil droplets above a target diameter

of 150 µm. For large flows, an array of several channels will be used in parallel. Figure 2 shows the typical internal structure of an API separator. The flow is passed through a coarse screen and then through flow baffles into the channels. Heavy solids settle to the bottom and are moved by a scraper to a sludge hopper. The oil droplets float to the surface. The length of the API separator is sufficient for most droplets above the target diameter to reach the surface. A simple underflow baffle/overflow weir at the outlet retains the oil on the surface. Oil is removed via a skimming trough located on the surface just upstream of the underflow baffle.

Figure 2 – API Separator

The advantage of the API separator is that it can deal with surges of flow with high solids and oil loading and is therefore ideal for treating oily water from tankage and process areas. The free oil content at the outlet is typically about 150 mg/l. Extending the length of the unit has little effect on performance due to residual turbulence and rapidly diminishing rise rates for smaller droplets. Additional oil removal is usually recommended prior to biological treatment.

Corrugated Plate Interceptor (CPI) A simple gravity separator can be intensified by adding submerged tilted plates. The plates are typically spaced 25 to 50 mm apart and set at an angle of 30 to 45 degrees from the horizontal. The design ensures that the flow between the plates is laminar. The oil removal efficiency of the plates can be further improved by corrugating the plates to cause oil droplets to touch and coalesce. The CPI was developed by Shell and can reliably remove oil droplets down to about 60 µm and achieve free oil concentrations of around 50 mg/l through the plate pack. Figure 3 shows the typical internal structure of a CPI with a grit trap, plate pack and oil skimmer.

Figure 3 – Corrugated Plate Interceptor

4

In a CPI, solids settle down towards the upper surface of a plate whilst lower-density oils float up towards the underside of the plate above. The solids slowly migrate down the plate surface and fall into the sludge hopper. The oil droplets rise into the top of the corrugations where they coalesce with other droplets. This increases the droplet size and increases the rise rate once they break out into the water above the plates and helps to ensure that the oil floats to the surface of the separator. Oil collected at the surface is removed by a skimmer. The underflow baffle is arranged so that all of the flow passes through the plate pack.

Dissolved Air Flotation Dissolved Air Flotation (DAF) enhances removal by attaching fine air bubbles to oil droplets and fine solids particulates. Figure 4 shows a typical flow scheme for a DAF unit. In a DAF unit the feed is conditioned by the sequential addition of a coagulant and flocculant to form flocs. As the flocs enter the main tank a recycle of treated effluent which has been saturated with air at elevated pressure is mixed with the conditioned feed. As the recycle is depressurised small bubbles form and attach to the flocs causing a floating effect. The flocs float to the surface and are removed by a scraper. The recycle flow is typically about 30% of the influent flow.

Figure 4 – Dissolved Air Flotation

The DAF will remove free oil and any colloidal oil which coagulates, and can typically achieve less than 20 mg/l of free oil which is ideal for biotreatment.

5

Comparison of Oil Removal Processes

Table1 summarises the typical footprints required to treat a wastewater flow of 1,000 m³/h.

Table 1 – Comparison of Oil Removal Processes to treat a wastewater flow 1,000 m³/h

Unit Main Tank Total

API separator

1,600 m² (8 units, each 40 m x 4 m) 3,000 m² (60 m x 50 m)

CPI 225 m² (10 m x 22.5 m) 1,250 m² (50 m x 25 m)

DAF 1,000 m² (20 m x 50 m) 2,500 m² (50 m x 50 m)

The CPI is the most compact but needs to be considered in the light of the operational advantages and disadvantages of these options (see Table 2).

Table 2 – Comparison of Constraints for Oil Removal Processes

Unit Advantages Disadvantages Mitigation

API separator

Huge capacity to absorb spills. Significant oil recovery.

Large internal submerged scraper systems are difficult to access for maintenance.

Design with multiple sub-units allows one to be taken off-line

(with careful flow management).

CPI

No submerged moving parts. Minimal energy

requirements. Easy to cover for odour control. Significant oil recovery.

Can be overwhelmed by spills. Risk of reduced efficiency due to blockage with accumulations of mud, asphaltenes and wax accumulations.

Upstream tank can intercept spills but needs mud

management. Design with multiple sub-units allows one to be taken off-line for cleaning (with careful flow

management).

DAF High quality effluent. Removal of

substantial fractions of colloidal oil.

Performs best with a pre-treatment stage to remove gross free oil (settlement tank, API or CPI). Consumes power for saturator recycle. More maintenance effort. Requires chemical dosing

systems.

Ensure adequate pre-treatment, so only fine and colloidal oil needs to be removed in the DAF.

Note - there will be many other site-specific advantages and disadvantages to consider and a range of further options which might also need to be considered with care. Biological Treatment The principal biological systems used in industrial process plants are:

• Conventional Activated Sludge • Membrane Bioreactor (MBR) • Trickling Filters

Other technologies which have been used in specific circumstances include moving-bed bioreactors (MBBR), rotating biological contactors (RBC) and sequencing batch reactors (SBR). Schemes in other industries have, after successful pilot test work, used submerged aerobic filters and anaerobic systems.

6

Conventional Activated Sludge

The conventional activated sludge process uses an aerated suspension of bacteria in the effluent to remove organic materials. The bacteria naturally form into flocs which are settled in a clarifier and recycled to the reactor inlet. The bacteria typically grow rapidly and the excess has to be removed to keep the reactor conditions optimal.

Sludge quality and sludge flux in the clarifiers is a consideration for designers and an on-going concern for operators. Industrial wastewater plant applications, especially where spent caustic is being treated, tend to suffer upsets which can cause solids loss from the clarifiers. If high-quality water is required, e.g. as feed to a water recovery stage, then a multi-media filter can be used to remove the solids. However, for systems where reliability is critical, experience suggests that a Bio-DAF unit should be included to protect the multi-media filters from overloading during upsets (see Figure 5).

Figure 5 – Conventional Treatment

Membrane Bioreactor (MBR)

MBRs use the same basic biological process as conventional activated sludge systems. However the solids retention is achieved by use of a membrane system. As this process is not constrained by sludge flux limits in clarifiers a higher sludge inventory can be maintained in the reactor. Typically, this is 10 g/l rather than the 3.5 g/l limit in the conventional process. The process does require a slightly lower loading per unit mass of sludge but the reactor is typically about 60% of the volume of a conventional system. The big advantage is that clarifiers, Bio-DAF and multi-media filtration are not required and are replaced by a much smaller membrane system (see Figure 6).

Figure 6 – Membrane Bioreactors

However, membranes are susceptible to fouling and require cleaning systems. Experience with MBR systems is improving all the time but in the absence of relevant prior experience, a treatability trial is recommended.

7

Trickling Filters

In a trickling filter, effluent is distributed across a deep bed of random-packed media and trickles downwards. Bacteria grow in a thin film on the media and consume the organics. A portion of the flow is usually recirculated to ensure that the bacteria remain wetted and /or to dilute the strength of the feed. Once the film of bacteria on the media gets too thick it breaks away or ‘sloughs off’ and is removed in a clarifier or ‘humus tank’. Air flows through the bed, typically by natural circulation but might require a fan (see Figure 7).

Figure 7 – Trickling Filters

Trickling filters are often cost- effective when used as high-rate ‘roughing filters’ to provide partial treatment.

The advantage of the roughing filter is that it removes a proportion of the organic load. It can be up to 10m high and so is taller than a typical activated sludge-based system, which is conventionally around 6-7m high. This extra height and the higher loading per cubic metre results in a smaller footprint. The power consumption is relatively low with just recirculation pumps required unless the design requires additional air from a fan or blower. The constraints are that it can only be used to remove about 60% of the load and solids removal is still required. The effluent will require further treatment in a conventional treatment plant to comply with many discharge consents. In these systems partially treated volatiles can be air stripped into the off-gas. This may mean that the air flowing away from the filter may require abatement to meet local air emission standards.

Comparison of Biological Processes Table 3 summarises the footprint and treatment intensity of various processes. The most intense is the roughing filter; the MBR is next with the conventional process taking up the largest footprint. The treatment intensity is shown in parentheses in the table. The design load of attached growth

systems is often expressed as a load per m³ of packing/media. One typical design parameter for an

activated sludge-based system uses the Food to Micro-organism (F/M) Ratio (kg COD*/kg bacteria

per day). Clarifiers are often designed on a flow per unit of surface area m³/m² per hour), so long as

the sludge-flux limits are not exceeded. *COD = Chemical Oxygen Demand

Table 3 – Comparison of Biological Processes

Unit Reactor Solids Removal

Roughing Filter 10,000 m³ (4 kg COD/m³.d) N/A

Conventional Activated Sludge 25,000 m³ (0.45 F/M COD) 1,700 m² (0.6 m³/m².h)

MBR 16,000 m³ (0.25 F/M COD) 1,250 m²

8

The roughing filter is the most compact process but needs to be considered in the context of the operational advantages and disadvantages of these options (see Table 4).

Table 4 –Constraints of Biological Processes

Unit Advantages Disadvantages Mitigation

Roughing Filter Robust. Low operating costs.

Channelling (by-passing of some packing).

Consider structured packing or improved distribution.

Conventional Activated Sludge

Robust and well proven.

Operators will be concerned about clarifier upsets leading to problems in the water upgrading stage.

Provision of good equalisation and

experienced operators.

MBR High quality treatment. No concern about clarifier upsets.

Risk of contamination of membranes. Membrane cleaning regime required.

Provision of thorough oily water equalisation and treatment system.

Note - there will be many other site specific advantages and disadvantages to consider and a range of further options which might also need to be carefully considered. The Intensification Challenge A ZD system, especially one that is integrated with the raw water supply, must work reliably under all credible operating conditions. Often this knowledge is gained once the process has been installed and the operators learn the practical limitations of a process. At first sight, the ideal process stage is usually the lowest cost and most intensified option. However, selecting the best option requires the operational advantages and disadvantages, especially reliability, to be carefully considered in view of the site-specific constraints. This requires the designer to produce a realistic model, or mass balance, of the flows and pollution loads to the various stages. In addition to conventional bulk design concerns such as organics, oil and suspended solids, salts and some individual components will usually need careful modelling. This requires a detailed knowledge of the composition of the effluent streams and an understanding of how the various treatment stages deal with each component. Based on many years’ experience, Foster Wheeler has developed a tried and tested in-house model, WM-SET (Water Management – System Evaluation Tool), which allows realistic modelling of complex effluent treatment and water recovery systems. Example Greenfield Refinery – Asia During the design of a new greenfield refinery in Asia, the national regulator ruled that the scheme would be required to achieve substantial water recovery and meet a tight daily flow limit for the

discharge to sea. Figure 8 shows a block flow diagram of the system adopted.

9

Figure 8 – Greenfield Refinery - Asia

Conventional effluent treatment would be expected to meet IFC and local standards. However, the additional requirement to recover clean water causes the residual contaminants to be concentrated in the reject streams. The further constraint of the flow limit causes the concentration of Chemical Oxygen Demand (COD) in the final reject stream from the water upgrading stage to exceed the IFC and local standards. As a result, the scheme required a stage to remove the COD residuals. The oil removal system at this site will consist of a mixture of API separation, tilted plate interceptors and DAF in series. This was driven in part by the operating company’s experience in their other refineries, thus concerns about operability prevailed over the opportunities for intensification. The biological treatment system uses a trickling filter as the first stage to minimise power requirements, capital cost and footprint and uses activated sludge as the second stage to maximise the removal of biodegradable components which would otherwise foul the water upgrading system. In this case, MBR was rejected by the operating company due to lack of experience and concerns about the potential impact of upsets in the oil removal system. An evaporator crystalliser was considered for removal of the residuals but rejected due to project-specific capital cost constraints in favour of an option with low capital cost but higher operating costs. Foster Wheeler was recently involved in another project in Asia with the expansion of a petrochemical plant on a site with very limited space for effluent treatment. The operating company pre-empted the space problem and commissioned a treatability trial at an early stage which demonstrated the viability of MBR. Concerns about upsets in the oil removal systems will be managed through instrumentation and operator training. Other projects have used MBBR to intensify the inlet end of a biological reactor and to enable an upgrade of an existing facility to allow nitrification to occur.

10

Should we use intense processes? A steadily increasing number of the recent projects that Foster Wheeler has developed for clients required us to consider intensification of effluent treatment processes and water recovery, either as upgrades or for new facilities. Successful implementation will typically require an iterative process involving but not limited to the following steps:

• identify effluents and water re-use opportunities/requirements • identify treatment steps and potential technologies and create block flow diagrams • identify opportunities for intensification and project constraints • seek advice on the performance and reliability of the intensified technologies • review the advantages, disadvantages, operating experience and select most effective technologies

• visit plants where the technology has been efficiently implemented • if suitable operational sites cannot be identified, undertake pilot testing with a vendor such that a guarantee can be obtained

• create flow sheet models for viable options and examine credible extreme scenarios • develop the detail and cost estimates to allow selection of the preferred scheme • produce a detailed design and implement with careful attention to guarantees for performance and availability

• ensure that operations and maintenance teams are properly prepared and trained prior to start-up

Conclusion Industrial sites are becoming ever more complex and to fit all the required treatment into a tight space, more intense processes are attractive. As engineers and designers we must however assess the practicality of the intensified schemes considered. To achieve this, we need to understand the constraints of the intensified process. If we can design the overall process to eliminate the constraints then the intense process can be used effectively. If the constraints cannot be adequately managed the process should not be used in that system. Foster Wheeler designs and constructs leading-edge processing facilities for the upstream oil and gas, LNG and gas-to-liquids, refining, chemicals and petrochemicals, power, mining and metals, environmental, pharmaceuticals, biotechnology and healthcare industries. Foster Wheeler works with this wide range of clients from the earliest stages of their projects to develop effluent treatment solutions which meet our clients’ objectives. ©2011 Foster Wheeler. All rights reserved.